Process a Perl qr RegEx string in reverse direction - regex

Question

I have a series of qr// RegEx string patterns to match URLs fed to my site. For example, qr#^/safari/article/([0-9]+)\.html(\?(.*))?$#. This string would match a path from a URL such as /safari/article/299.html?parameter=1. I have a separate subroutine where I can create URLs to link to different parts of the program. It occurred to me that it would be nice if that latter part could somehow use those aforementioned patterns I have already written -- it would reduce the likelihood of error if both the way URLs were generated and the way they are later processed came from the same set of patterns.
When a user comes to the site, my program takes the URL given to the server, runs it against strings like the one above and it will output $1 and $2 with the patterns it finds (e.g. "299" and "parameter=1," the two parameters for loading a page). In essence, now I'd like to do that in reverse and somehow provide $1 and $2 and feed them against that qr// string to create a new path output (say, I'd set $1 to "300" and $2 to "parameter=2," somehow merge that against the qr// string and get the output /safari/article/300.html?parameter=2.
Is there a simple way to do that sort of "reverse regex"? It seems like one way to do it would simply be to do a regex pattern match against those two parenthetical patterns, but that somehow feels sloppy to me. Is there a cleaner way?
EDIT: Part of the reason for storing the patterns in RegEx is that they all get thrown into a multidimensional array for later processing that can help figure out what module should be called. Here's a couple of sample items:
[
{ function => 'article', pattern => qr#^/safari/article/([0-9]+)\.html(\?(.*))?$#, weight => 106 },
{ function => 'topCommentedPosts', pattern => qr#^/safari/top\.html$#, weight => 100 }
]

Answer 1

I'm not sure I understand exactly what you want to achieve. The following works, but going this ways seems rather fragile and dangerous. Why do you need to generate the paths, anyway?
#!/usr/bin/perl
use warnings;
use strict;
use feature qw{ say };
my $TEMPLATE = '/safari/article/$1.html?$2';
sub generate {
my (#replacements) = #_;
return $TEMPLATE =~ s/\$([12])/$replacements[$1-1]/gr
}
sub match {
my ($string) = #_;
my $regex = "$TEMPLATE";
$regex =~ s/([?.])/\\$1/g;
$regex =~ s/\$[0-9]+/(.*)/g;
return $string =~ /$regex/
}
use Test::More;
is generate(300, 'parameter=2'), '/safari/article/300.html?parameter=2';
is_deeply [match('/safari/article/299.html?parameter=1')], [299, 'parameter=1'];
done_testing();